AI News, BOOK REVIEW: Lessons from the NSA: Know Your Assets artificial intelligence

The Cyberlaw Podcast

Will Turner explained that the TurnKey No-Action Letter was most useful for parties interested in structuring a private, permissioned, centralized blockchain, but believes the guidance in the Framework would allow for alternative structures.

Blockstack is seeking to be a Tier 2 issuer, meaning they can raise up to $50 million in 12 months, which comes with heightened disclosure obligations and requires audited financials.

While they seek to raise capital as a security today, their ultimate goal – and a central risk factor in their offering circular – is to achieve the requisite level of decentralization such that they no longer would meet the definition of a security.

Meanwhile, in Congress, the recently reintroduced Token Taxonomy Act of 2019 would exempt a newly defined category of digital tokens from the definition of a security, as well as provide some clarity on tax issues for cryptocurrency users and exchanges.

Turning to the interview, Jeff describes how he co-founded Global Digital Finance (GDF), along with other co-founders in Europe, Asia and the United States, in order to address the lack of international standards surrounding the blockchain industry—or even a general consensus of terminology.

Jeff describes how GDF has a number of working groups focused on developing high-level principles and standards on a range of topics, including stablecoins, custody, tax and security tokens.

As markets mature, it is increasingly important to have the necessary post-trade infrastructure, and he is committed to offering services that recognize the novel features and efficiencies around these new technologies.

Home Page for Professor Michael Kearns, University of Pennsylvania

the science of designing algorithms that embed social values like privacy and fairness;

Over the course of a generation, algorithms have gone from mathematical abstractions to powerful mediators of daily life.

Algorithms have made our lives more efficient, more entertaining, and, sometimes, better informed.

At the same time, complex algorithms are increasingly violating the basic rights of individual citizens.

Allegedly anonymized datasets routinely leak our most sensitive personal information;

statistical models for everything from mortgages to college admissions reflect racial and gender bias.

Meanwhile, users manipulate algorithms to 'game' search engines, spam filters, online reviewing services, and navigation apps. Understanding

and improving the science behind the algorithms that run our lives is rapidly becoming one of the most pressing issues of this century.

Traditional fixes, such as laws, regulations and watchdog groups, have proven woefully inadequate.

Reporting from the cutting edge of scientific research, The Ethical Algorithm offers a new approach: a set of principled solutions based on the emerging and exciting science of socially aware algorithm design.

Michael Kearns and Aaron Roth explain how we can better embed human principles into machine code - without halting the advance of data-driven scientific exploration.

Weaving together innovative research with stories of citizens, scientists, and activists on the front lines, The Ethical Algorithm offers a compelling vision for a future, one in which we can better protect humans from the unintended impacts of algorithms while continuing to inspire wondrous advances in technology.

Understanding and improving the science behind the algorithms that run our lives is rapidly becoming one of the most pressing issues of this century.

Traditional fixes, such as laws, regulations and watchdog groups, have proven woefully inadequate.

Reporting from the cutting edge of scientific research, The Ethical Algorithm offers a new approach: a set of principled solutions based on the emerging and exciting science of socially aware algorithm design.

Michael Kearns and Aaron Roth explain how we can better embed human principles into machine code - without halting the advance of data-driven scientific exploration.

Weaving together innovative research with stories of citizens, scientists, and activists on the front lines, The Ethical Algorithm offers a compelling vision for a future, one in which we can better protect humans from the unintended impacts of algorithms while continuing to inspire wondrous advances in technology.

Here are links to some media related to the book:

Most of this site is organized as a single flat html file.

The links below let you navigate directly to the various subsections.

My research interests include topics in machine

learning, algorithmic game theory and microeconomics, computational

often examine problems in these areas using methods and models from theoretical computer

also often participate in empirical and experimental projects, including

applications of machine learning to problems in algorithmic trading and quantitative finance, and

strategic and economic interaction in social networks.

QUICK LINKS Videos of some recent talks:

Panel discussion on privacy policy and regulation (AT&T

Networked and Social Systems Engineering (NETS) Program Penn undergraduate course

Since 2002 I have been a professor in the Computer and Information Science Department at the University of Pennsylvania, where I hold the National Center Chair. I

am the faculty founder and former director of

Networked and Social Systems Engineering (NETS) Program, whose current directors are Andreas Haeberlen and Aaron

whom I have also collaborated on a number of papers on algorithmic trading ), I

have worked extensively in quantitative and algorithmic trading on

often serve as an advisor to technology companies and venture capital firms.

occasionally invest in early-stage technology startups. I

FINRA.I occasionally serve as an expert witness or consultant on technology-related legal and regulatory

spent the decade 1991-2001 in machine learning and AI research at AT&T Bell Labs.

was the head of the AI department, which conducted a broad range of systems and foundational AI work; I

also served briefly as the head of the Secure Systems Research department.

The AI department boasted terrific colleagues and friends that included Charles

to my time as its head, the AI department was shaped by the efforts

Kautz (who led the department before heading to the

leading the AI group, I was a member of the closely related Machine

spent 2001 as CTO of the European venture capital firm Syntek Capital, and

trading, technology and asset management firm based in NYC. From

led a quantitative portfolio management team with Yuriy Nevmyvaka at Engineers

led a quantitative trading team at Bank

of America in New York City, working on both proprietary and algorithmic trading strategies within

the Spring of 2002 through May 2007, I was first a consultant to, and later the head of, a quant prop trading team within the Equity Strategies group of Lehman Brothers in New York City.

did my undergraduate studies at the University of California at Berkeley in math and computer science, graduating in 1985.

Publications below for more information), and Les Valiant was my (superb) advisor.

1991 I joined the research staff of AT&T Bell Labs, and later the Penn faculty (see professional bio above). Alongside

my formal education, I was strongly influenced by being born into an academic family, including my

I have also served on the program committees of NIPS, AAAI, IJCAI, COLT, UAI, ICML, STOC, FOCS, and a variety of other acryonyms.

I am a member of the NIPS Foundation, and was formerly on the steering committee for the Snowbird Conference on Learning (RIP).

2002-2008 I was a member, vice chair and chair of DARPA's

Current (alphabetical):

Neel (jointly advised with Aaron Roth )

Alumni (reverse chronological):

TEACHING AND TUTORIAL MATERIAL In Spring 2019, I piloted a new undergraduate course on

slides for my STOC 2012 tutorial on Algorithmic Trading and Computational Finance

Web page for the graduate seminar course

Web page for the graduate seminar

This was a joint course between CIS and Wharton (listed as CIS 620 and Wharton OPIM 952).

Course web page for CIS 620: Advanced Topics in AI (Spring 2002)

1999 Bellairs Institute Workshop Theoretical Issues in Probabilistic Artificial Intelligence (FOCS 98 Tutorial) [PDF]

article on the market for consumer data and related privacy concerns, October

Marketplace interview on presidential tweets, market volatility and algorithms (roughly the 2 minute mark), August

Skeptic' podcast with Kyle Polich on machine learning, computational complexity, game theory, trading, fairness etc. November

Magazine (Santa Fe New Mexican) article about SFI lecture on machine learning and social norms, &nbsp April

Bloomberg News article on HFT and hybrid quant funds, March 2014

NPR text and audio on Coursera, online education, and Penn, October 2012

Fiscal Times article on machine learning and technology in trading, March 2011, Wired

News article on light speed propagation delays in trading, October 2010 Economist

Atlantic article on HFT 'crop circles', August 2010 Nature News article on 'distributed thinking', August 2010 Wall Street Journal article on machine learning in quant trading, July 2010 and a related interview on CNBC New Scientist article on 'Why Facebook friends are worth keeping', July 2010;

Philadelphia Business Journal article on the MKSE program and Networked Life, October 2009 Discussion of behavioral social network experiments in Christakis and Fowler's 'Connected'

Philadelphia Inquirer article on networked voting experiments, March 2009 Science Daily article on networked voting experiments, February 2009

natural language processing for algorithmic trading, September 2007

Philadelphia Inquirer article on network science and NSA link analysis, May 2006

Chicago Tribune article on privacy in blogs and social networks, November 2005

Philadelphia Inquirer article on ISTAR, interdependent security, and games on networks, January 2003 Washington Post article on web-based chatterbots, September 2002 New Scientist article on the Cobot spoken dialogue system, August 2002 Tornado Insider article on DDoS attacks, January 2002 [Cover] Tornado Insider article on biometric security, January 2002 Audio of COMNET panel 'Staving Off Denial-of-Service Attacks and Detecting Malicious Code' Tornado Insider article on natural language technology, September 2001 Tornado Insider article on robotics, July 2001 Il Sole 24 Ore profile, June 2001 [English Translation] Corriere Della Sera profile, May 2001 [English Translation] Associated Press article on software robots, February 2001 New York Times article on TAC, August 2000 New York Times on Cobot, February 2000 TIME Digital Magazine (now Time On) on Cobot, May 2000 Washington Post article on Cobot, December 2000 New York Times article on boosting, August 1999

the science of designing algorithms that embed social values like privacy and fairness.

Berkeley, this MIT Press publication is intended to be an intuitive but precise treatment of some interesting and fundamental topics in computational learning theory.

The level is appropriate for graduate students and researchers in machine learning, artificial intelligence, neural networks, and theoretical computer science.

The link above is to the MIT Press page that provides a brief description of the book and ordering information.

If you have the book and have any questions or comments, please click here to send me mail.

revision of my doctoral dissertation was published by the MIT Press as part of the ACM

it is now out of print, I am making it available for downloading below.


What follows is a listing of (almost) all of my research papers in (approximately) reverse chronological

as per the honorable tradition of the theoretical computer science community, on almost all of the papers

below that are primarily mathematical in content, authors are listed alphabetically.

In addition to the list below, you can also look at my page on Google

to do a pretty good job of finding those publications that appeared in mainstream CS venues (though not others), and

(This paper is an expanded version of the Goyal-Kearns STOC 2012 paper, and

is a corrected version posted Oct 4 2006. This

version corrects errors in the section of experimental results published

With Y.

With S.

COLT 2005.

With S.

Proceedings of NIPS 2004.

With S.

Proceedings of COLT 2004.

With S.

Proceedings of the ACM Conference on Electronic Commerce (EC), 2004.

With L.

NIPS 2003.

With S.

ACM Conference on Electronic Commerce (EC), 2003.

With L.

IEEE Intelligent Systems, Nov/Dec 2003.

With S.

ICML 2003.

With S.

Journal of Artificial Intelligence Research, 2002.

With L.

Proceedings of NIPS 2002.

With Y.

Proceedings of UAI 2002.

With E.

Proceedings of NIPS 2002.

Proceedings of AAAI 2002.

The main result of the paper above --- an efficient algorithm claimed to find a single

Nash equilibrium in tree graphical games --- is unfortunately

This was discovered and discussed in the very nice paper by Elkind, Goldberg and Goldberg, which can be found here.

The problem of efficiently computing an exact Nash equilibrium in trees remains open (though EG&G demonstrate that no two-pass algorithm can suffice).

Nash algorithm from the K., Littman, Singh UAI 2001 paper is unaffected by these developments, as is its NashProp generalization in the Ortiz and K.

Journal of Artificial Intelligence Research .

Earlier version in Proceedings of Agents 2001.

[PDF] New York Times article on TAC

Computational Learning Theory, 2000,

pages 142--147.

via Reusable Trajectories.

With Y.

MIT Press, 2000.

428-456, 2000. Earlier version in Proceedings of the 12th Annual Workshop on Computational Learning Theory.

Appeared in COLING 2000.

Appeared in ICML 2000.

MIT Press, 2000.

at the Dialogue Level.


pages 309-316.

Planning in Large Markov Decision Processes.

With Y.

Q-Learning and Indirect Methods.

With S.

Advances in Neural Information Processing Systems 11, The MIT Press, 1999, pages 996--1002.

with L.

Advances in Neural Information Processing Systems 11, The MIT Press, 1999, pages 260--266.

with Rates

of Convergence.

Structure from Sample Data in Noisy-OR Networks.

With Y.

Proceedings of the 15th International Conference

on Machine Learning, pp.

on Machine Learning, 1998, Morgan Kaufmann, pages 269--277.

Analysis of Hard and Soft Assignment Methods for Clustering.

with Y.

Thirteenth Conference on Uncertainty in Artificial Intelligence,


Abstract accompanying invited talk given at AAAI '96, Portland, Oregon, August 1996.

pages 109-128.

Validation Using

the Approximation and Estimation Rates, with Consequences for the

Training-Test Split.

Earlier version in Advances in Neural Information Processing Systems 8, The MIT Press, pages 183--189, 1996.

of Model Selection Methods.

Statistical Mechanics Theory of Learning Curves for

the Model Selection Problem.

Neural Networks: The Statistical Mechanics

Perspective, pp.

Artificial Intelligence, 74(1), pages 129-145, 1995.



26th Annual ACM Symposium on the Theory of Computing,


AAAI 1992.


missing references, bib file got corrupted)

Earlier version in Proceedings of the 1988 Workshop on Computational Learning Theory, pp.

Unpublished manuscript, 1988.


19th ACM Symposium on the Theory of Computing, pp.

1987, ACM Press.

With M.

41(6), pp.

19th ACM Symposium on the Theory of Computing, pp.

1987, ACM Press.

Concept Learning.

Proceedings of the Fourth International Conference

on Machine Learning, pp.

Daily briefing.

Cybersecurity: This password-stealing hacking campaign is targeting governments around the world (ZDNet) Researchers uncover a phishing campaign attempting to steal login credentials from government departments across North America, Europe and Asia - and nobody knows who is behind it Over 100 Fake Sites Spotted in Government Phishing Campaign (Infosecurity Magazine) Over 100 Fake Sites Spotted in Government Phishing Campaign.

look at the recent BuleHero botnet payload (Zscaler) Increase in adoption of lateral propagation modules in malware, deep insight into BuleHero botnet, lateral movement, Gh0st RAT, XMRig Miner New Echobot Variant Exploits 77 Remote Code Execution Flaws (BleepingComputer) The Echobot botnet is still after the low hanging fruit as a new variant has been spotted with an increased number of exploits that target unpatched devices, IoT for the most part.

Michigan family says hacker was talking to them through security system in their home (WWMT) Some people put security cameras outside and inside their homes to protect themselves, but if you're not careful, tech experts said those cameras can actually give hackers an inside look.

Global New Account Fraud Increased 28% in 2019, According to Jumio Holiday Fraud Report (BusinessWire) New data from Jumio, the leading AI-powered trusted identity as a service provider, reveals that new account fraud increased 27.8% worldwide YTD in 20 Consumers not willing to compromise when it comes to IoT security (Help Net Security) Nearly three quarters of consumers expect manufacturers of connected IoT devices to protect their devices from hacks, according to Karamba Security.

42 More Cybersecurity Predictions For 2020 (Forbes) From disrupting elections to targeted ransomware to privacy regulations to deepfakes and malevolent AI, here are additional 42 predictions from senior cybersecurity executives Why the election result gives UK tech sector a silver lining in the Brexit cloud (The Telegraph) In a general election in which parties delivered their political messaging with memes and spent an unprecedented amount on digital ads, technology has unquestionably been front and centre of a bitterly-fought contest that looks set to see the Conservatives emerge victorious.

China’s Huawei may need two to three years to recover from U.S. trade ban, CEO says (Washington Post) In an interview at Huawei headquarters in southern China, Ren Zhengfei said the U.S. ban has had a “pretty big impact” on the company, forcing it to scramble to redesign products to try to eliminate U.S. parts.

Mimecast UK channel boss on 'luring' Symantec customers (ChannelWeb) Tom Corrigan opens up about battling the political and economic events affecting the email security vendor's UK growth Facebook Won’t Change Web Tracking in Response to California Privacy Law (Wall Street Journal) Facebook maintains that it doesn’t need to make changes to its web-tracking services in order to comply with California’s new privacy law, setting up a potential early clash over the measure set to take effect Jan. 1.

Facebook Pledges $130 Million to Fund Content Oversight Board as It Hits Delays (Variety) Facebook announced an initial commitment of $130 million to launch its new Oversight Board, which is designed as a way for users to appeal the social giant’s content decisions, but the entity… Bugs, bounties, and cherry browns (GitLab) Cheers, our bug bounty program is celebrating one year!

GDIT Wins Army IT, Cyber Support Contract (WashingtonExec) General Dynamics Information Technology has won an estimated $118 million contract to support the critical information network that is key to Army Cyber Work podcast write-up: How to become a cybersecurity analyst (Infosec Resources) Introduction This article will explore some interesting details from an episode of Infosec’s information security career podcast, Cyber Work.

This Fidelis Cybersecurity Appoints Former NSA CISO as the Company's Chief Information Security Officer (IT Business Net) Cybersecurity expert Chris Kubic joins Fidelis to provide customers with cybersecurity best practices, assist with product strategy and lead internal security Lastline Unveils Defender 9.0 to Enhance Public Cloud Workload Security (PR Newswire) Lastline®, the leader in AI-powered network detection and response, today announced the release of Lastline Defender 9.0 to secure public cloud...

As a result, the business … FireEye rolls out threat intelligence platform for industrial systems (Security Brief) Now industrial control systems (ICS), operational technology (OT), internet of things devices, and other equipment used to manage interconnected physical processes, can be secured from cyber threats.

For journalists, policy makers, and concerned consumers, this paper will demystify the fundamentals of third-party tracking, explain the scope of the problem, and suggest ways for users and legislation to fight back against the status quo.

Facebook asks users for help in spotting deepfakes (Computing) The winner of Facebook's challenge will take home a $500,000 prize AI expert calls for end to UK use of ‘racially biased’ algorithms (the Guardian) Prof Noel Sharkey says systems so infected with biases they cannot be trusted Air Force to Decide Which Satellite to Offer for Test at Defcon Hacker Conference (Avionics) The United States Air Force plans to choose on Dec. 13 which satellite it will offer to a service-selected group of hackers to try to infiltrate at next August’s Defcon 2020 conference in Las Vegas, an Air Force official said on Dec. 11.

It also would create a two-year pilot Senate bill would give DHS cyber agency subpoena powers (TheHill) Two senators unveiled bipartisan legislation on Thursday that would give the Department of Homeland Security’s (DHS) cyber agency the ability to subpoena internet service providers to increase transparency about cyber vulnerabilities.

New bill gives this Homeland Security agency authority to get ISP info (Fifth Domain) Senators, however, say the authorities are Defense bill limits commerce secretary's ability to remove Huawei from Entity List (Inside Defense) The fiscal year 2020 defense authorization conference bill includes language requiring congressional notification before the commerce secretary can remove Huawei from the 'Entity List' restricting sales of U.S. components to the Chinese telecom, a procedural hurdle added to the bill amid congressional concerns over the Trump administration's commitment to maintaining a hard line on the cyber threat from China.

The goal is just one of several provisions aimed at… The Defense Department Says It Needs the Encryption the FBI Wants to Break (Vice) A bipartisan coalition of lawmakers this week worked overtime to vilify encryption, oblivious to the fact that weakening encryption standards will put the public, and the internet, at risk.

No new foreign military students allowed in the US until better checks, says DoD spokesman (Military Times) No new international military students will come to the United States for training until new screening procedures are in place, the Pentagon said Thursday in the wake of the deadly shooting last week by a Saudi Arabian aviation trainee at a Florida Navy base.

(C4ISRNET) A prominent group of national security thinkers is questioning if the Pentagon’s policy on developing autonomous weapons needs to be updated to more accurately reflects current technology and the greater role artificial intelligence is expected to play in future conflicts.

Exploring Quantum Computing In Asset Management

Christian Weedbrook recently joined Jean-Sebastien Cournoyer at the Trusted Insight Summit 2019. Weedbrook is the founder & CEO of Xanadu, a photonic ...

Alex Tapscott: "Blockchain Revolution" | Talks at Google

Talks at Google in London were delighted to welcome Alex Tapscott to talk about his book Blockchain Revolution, looking at how the technology behind Bitcoin ...

Professor Dawn Song from Oasis Labs on her new privacy blockchain

Professor Dawn Song from Oasis Labs, a Chainfund Capital portfolio company, talks about crypto on CNBC Crypto Trader. They are building a privacy-first, ...

Artificial Intelligence And Human Rights | SkollWF 2019

As the uses of artificial intelligence grow exponentially each year, there is a possible future where AI exacerbates injustice, deepens societal divides, and ...

SOC 2030—SOCs Are Broken. Let’s Fix Them.

Kerry Matre, Head of Security Operations Strategy, Palo Alto Networks Is your SOC overwhelmed with alerts and threats? Cyber-adversaries are wielding tools ...

Security Operations 2018: What Is Working? What Is Not.

Kerry Matre, Security Operations Strategist, Palo Alto Networks Security operations have had to evolve over the last four decades to keep up with the advancing ...

DevOps Tutorial for Beginners | Learn DevOps in 7 Hours - Full Course | DevOps Training | Edureka

Edureka DevOps Training (Use Code: YOUTUBE20): ** This Edureka DevOps Tutorial for Beginners will ..